Computationally Optimizing the Directed Evolution of Proteins

Author: Voigt, Christopher Ashby

Year: 2003

Degree: Dissertation (Ph.D.)

Advisors: Wang, Zhen-Gang; Arnold, Frances Hamilton; Mayo, Stephen L.

Committee Members: Rees, Douglas C.; Arnold, Frances Hamilton; Roberts, Richard W.; Mayo, Stephen L.; Fontana, Walter; Wang, Zhen-Gang

Option: Biochemistry and Molecular Biophysics

DOI: 10.7907/E4GF-EQ41

Abstract

Directed evolution has proven a successful strategy for protein engineering. To accelerate the discovery process, we have developed several computational methods to optimize the mutant libraries by targeting specific residues for mutagenesis, and subunits for recombination. In achieving this goal, a statistical model was first used to study the dynamics of directed evolution as a search algorithm. These simulations improved our understanding of the relationship between parameters describing the search space (e.g., interactions between amino acids) and experimental search parameters (e.g., mutation rate and library size). Based on these simulations, a more detailed model was used to calculate the structural tolerance of each residue to amino acid substitutions. Further, a computational model was developed to optimize recombination experiments, based on the three-dimensional structure. Together, these computational techniques represent a major step towards information-driven combinatorial protein design.

Files